1. Field of the Invention
The present invention relates to a three-dimensional image capturing device detecting information relating to a three-dimensional shape or topography and a two-dimensional image of a subject, and its recording medium.
2. Description of the Related Art
A three-dimensional measurement using a three-dimensional image capturing device of the related art may be classified into active systems radiating light, an electromagnetic wave, or a sonic wave to a subject and passive systems not emitting light etc. As active systems, time-of-flight, phase change, triangulation, moire contouring, interferometry, etc. are known. As passive systems, stereo vision, range-from-focus, and so on are known.
As a method of storing two-dimensional image information of the subject together with information of the distance to a subject obtained by one of the above three-dimensional measurement methods, there is the image format standard Exif used in digital still cameras (JEIDA standard). The distance to the subject is detected by using an auto focus mechanism etc., and the obtained distance measurement data is stored in a file management area while the two-dimensional image data (hereinafter referred to as the “image data”) is stored in the data area. This format, however, only comprises a single distance measurement data related to the subject. Therefore the data stored by this format cannot be used for three-dimensional image processing based on a three-dimensional shape or topography of the subject.
Further, the apparatus disclosed in Japanese Unexamined Patent Publication (Kokai) No. 8-317425 stores pairs of image data for the left eye and for the right eye in the stereo vision method in a single recording medium. In the file management area, it also stores the link information for identifying the pairs of image data. Storing the image data by the above method, however, requires two sets of image data to be stored in order to obtain single three-dimensional image information and therefore about two times the storage space is required compared with ordinary image data. Further, in order to calculate the distance from the apparatus to a certain point on the surface of the subject corresponding to a certain pixel of the above images, the corresponding relations between pixels in the right and left eye image should be detected. However, detecting the corresponding relations between pixels in the right and left eye image is not easy, so it is difficult to obtain the distance measurement data corresponding to each of the image pixels. Consequently, it is inconvenient in the three-dimensional image processing to make direct use of distance measurement data corresponding to each of the image pixels.